In this episode, Dave and Jamison answer these questions:

My parent organization has trust issues: we registered on a recent survey as one of the lowest across the bigger software org (thousands of employees).

There are two groups: functional, trustworthy people who get stuff done, and people who are behind, stuck, or just not working. Those struggling say they need better emotional support, but there is consistent, documented evidence that they cannot keep up.

I’m perpetually frustrated that there are only two or three people in an org of 30 who can effectively complete tasks and manage the insane workload. I am biased: those in whom I have no trust have repeatedly demonstrated that they cannot be trusted. I believe that the organization would be able to go faster without them.

Whats the right answer here? Should I start my own company to abandon this mess? Do we cut scope super aggressively to allow underperformers to be reasonable contributors?

One example: one of these contributors was walked through the process, given written documentation of the process, verbally confirmed an understanding of the process, and committed to starting that day. And then two days later identified they hadn’t started for those two days because they were blocked by something that was explicitly captured in the document and discussed in the recorded meeting. They do not raise this until they were asked for their progress multiple days beyond the critical start date.
Hi Dave and Jamison,

First, thank you for the podcast. As someone on the spectrum, it really helps me analyze social situations I struggle with.

My question: I work at a software company where management is pushing the use of LLMs for coding, and my team recently started using Spec-Driven Development. SDD often requires strong upfront planning, which has quietly split my small team between developers who plan well and those who get lost in “vibe-coding” loops.

One of those people is a close coworker I consider a friend (we hang out after work and have honest chats about everything). He’s not strong at planning, so I end up explaining each ticket to him in detail. But he’s really great in other ways: communicative, asks a lot of questions in refinement sessions that set a good example for others, does thorough handoffs with the QA team, always responsive in chat, always trying to help. He’s also one of the rare people who actually pay attention to alerts.

Since we started using SDD, the gap in our speed and output quality has become very visible. One of our tech directors noticed and asked me to teach him planning. We spent several days drawing schemas and working through small features together, and it was clearly painful for him. The TD also had his own sessions with him, but eventually gave up because my friend seemed so discouraged, and the TD decided to “stop the torture” and leave him alone. The hard part is that after all these teaching sessions, he actually seems even slower than before and also more discouraged and down.

Since then, our friendship has changed. He stopped talking to me outside work, and I think he now feels jealous or bothered by the difference in our performance. I’m okay with continuing to explain tickets and outlining detailed execution plans for him, but I worry that I’m keeping him in his comfort zone and not really helping his career. And I also miss how things used to be between us 💔

What should I do?

P.S. I’m not an LLM fan or enthusiast, I take a practical approach to using them. But I hate how they’re taking jobs away from devs who really love coding but are weak at planning or not interested in it. It doesn’t feel fair.